Coreceptor Scanning by the T Cell Receptor Provides a
Mechanism for T Cell Tolerance
The MIT Faculty has made this article openly available. Please share
how this access benefits you. Your story matters.
Stepanek, Ondrej, Arvind S. Prabhakar, Celine Osswald,
Carolyn G. King, Anna Bulek, Dieter Naeher, Marina BeaufilsHugot, et al. “Coreceptor Scanning by the T Cell Receptor
Provides a Mechanism for T Cell Tolerance.” Cell 159, no. 2
(October 2014): 333–345.
As Published
Author's final manuscript
Thu May 26 19:38:32 EDT 2016
Citable Link
Terms of Use
Creative Commons Attribution-NonCommercial-NoDerivs
Detailed Terms
NIH Public Access
Author Manuscript
Cell. Author manuscript; available in PMC 2015 October 09.
NIH-PA Author Manuscript
Published in final edited form as:
Cell. 2014 October 9; 159(2): 333–345. doi:10.1016/j.cell.2014.08.042.
Co-receptor scanning by the T-cell receptor provides a
mechanism for T cell tolerance
Ondrej Stepanek1,*, Arvind S. Prabhakar2, Celine Osswald1, Carolyn G. King1, Anna
Bulek3, Dieter Naeher1, Marina Beaufils-Hugot1, Michael L. Abanto1, Virginie Galati1,
Barbara Hausmann1, Rosemarie Lang1, David K. Cole3, Eric S. Huseby4, Andrew K.
Sewell3, Arup K. Chakraborty2,5,6,#, and Ed Palmer1,*
NIH-PA Author Manuscript
of Biomedicine and Nephrology, University Hospital Basel and University of Basel,
CH-4031 Basel, Switzerland 2Department of Chemical Engineering, Massachusetts Institute of
Technology, Cambridge, MA 02139 USA 3Institute of Infection and Immunity, Cardiff University
School of Medicine, Cardiff CF14 4XN, UK 4Department of Pathology, University of
Massachusetts Medical School, Worcester, MA 01655 USA 5Institute for Medical Engineering &
Science, Departments of Physics, Chemistry, and Biological Engineering, Massachusetts Institute
of Technology, Cambridge, MA 02139 USA 6Ragon Institute of MGH, MIT & Harvard, 400
Technology Square, Cambridge, MA 02139 USA
NIH-PA Author Manuscript
In the thymus, high affinity, self-reactive thymocytes are eliminated from the pool of developing T
cells, generating central tolerance. Here, we investigate how developing T cells measure selfantigen affinity. We show that very few CD4 or CD8 coreceptor molecules are coupled with the
signal-initiating kinase, Lck. To initiate signaling, an antigen engaged T cell receptor (TCR) scans
multiple coreceptor molecules to find one that is coupled to Lck. Coreceptor scanning is the first
and rate-limiting step in a kinetic proofreading chain of events that eventually leads to TCR
triggering and negative selection. MHCII-restricted TCRs require a shorter antigen dwell time
(~0.2s) to initiate negative selection compared to MHCI restricted TCRs (~0.9s) because more
CD4 coreceptors are Lck-loaded compared to CD8. Based on experimental data and mathematical
analysis, we generated a model (Lck come&stay/signal duration) that accurately predicts the
experimentally observed differences in antigen dwell-time thresholds used by MHCI- and MHCIIrestricted thymocytes to initiate negative selection and generate self-tolerance.
Correspondence: [email protected] (EP) or [email protected] (OS).
#Co-Senior Author
Author Contributions
OS, CO, CGK, MA, MB, DKC, AKS, DN, and EP designed experiments. OS, CGK, CO, MA, BH, VG, MB, RL performed
experiments and analyzed data. OS, ASP, and AKC generated the mathematical models. ESH provided mice and generated MHCII
monomers. OS and EP wrote the manuscript. All authors commented on the manuscript draft. AKC, AKS, and EP funded and
supervised the study.
The authors declare no conflict of interests.
Stepanek et al.
Page 2
NIH-PA Author Manuscript
T cells regulate adaptive immune responses to pathogens and tumors, but can also drive
autoimmune diseases. The T-cell antigen receptor (TCR) on conventional αβT cells
recognizes peptide fragments bound to class I or class II major histocompatibility complexes
(pMHCI or pMHCII). Each developing T cell expresses a unique TCR and generation of a
self-MHC restricted and self-tolerant T-cell repertoire results from a multistep selection
process in the thymus. Thymocytes expressing a TCR weakly reactive to the host's selfantigens receive a maturation signal to generate the functional T-cell repertoire in the
periphery (positive selection). In contrast, thymocytes with strongly self-reactive TCRs
receive a death signal (negative selection). A failure to prevent strongly self-reactive T cells
from entering the peripheral T-cell pool is one of the main causes of autoimmune diseases
(Yin et al., 2013). How thymocytes discriminate between positive and negative selecting
antigens in the thymus is incompletely understood. Another open question is how a
thymocyte balances the high sensitivity required to recognize just a few molecules of strong
antigens (Ebert et al., 2008; Peterson et al., 1999) with the selectivity needed to discriminate
between positive and negative selecting antigens even at relatively high densities (Daniels et
al., 2006; Naeher et al., 2007).
NIH-PA Author Manuscript
Engagement of a TCR by its cognate ligand leads to phosphorylation of TCR associated
ITAM-containing TCRζ and CD3 chains by a Src family kinase, Lck (Straus and Weiss,
1992). Antigen discrimination might already occur at this step, because positive selecting
antigens poorly induce phosphorylation of TCRζ chain (Kersh et al., 1998). Doubly
phosphorylated ITAMs recruit ZAP70, a kinase that is subsequently activated by a second
round of Lck-mediated phosphorylation (Straus and Weiss, 1993). ZAP70 relays the signal
downstream by phosphorylating LAT and SLP76 (Smith-Garvin et al., 2009).
NIH-PA Author Manuscript
The CD4 and CD8 coreceptors bind to MHCII and MHCI, respectively. It has been
suggested that the principal role of coreceptors is to enhance TCR signaling by delivering
Lck to an engaged TCR (Artyomov et al., 2010; Veillette et al., 1988). CD8 additionally
stabilizes TCR-pMHC interaction (Stone et al., 2009). Although signaling can be induced by
very strong agonists or anti-TCR antibodies in the absence of coreceptors (van der Merwe
and Dushek, 2011), CD4 or CD8 are required for signaling induced by most ligands (Kerry
et al., 2003; Vidal et al., 1999). Moreover, coreceptors are vitally important for selecting T
cells that recognize pMHCI and pMHCII antigens (Van Laethem et al., 2013). Along these
lines, increasing Lck coupling to CD8 enhances the efficiency of positive selection of
MHCI-restricted thymocytes (Erman et al., 2006).
The strength of a self-antigen-TCR interaction dictates whether a developing thymocyte
undergoes negative selection (Daniels et al., 2006; Hogquist et al., 1994; Williams et al.,
1999). The main parameters describing the interaction between a TCR and its ligand are
association rate (kon), dissociation rate (koff), and equilibrium dissociation constant (KD).
Whereas koff determines the median dwell time of the antigen-TCR interaction (τ1/2=ln2/
koff), kon (which is concentration dependent) determines the rate of TCR-pMHC complex
formation. KD (=koff/kon) indicates the concentration-dependent occupancy of the TCR
under equilibrium conditions. Although there are conflicting data whether kon, koff, KD, or
Cell. Author manuscript; available in PMC 2015 October 09.
Stepanek et al.
Page 3
NIH-PA Author Manuscript
aggregate dwell time better describes the biological response induced by particular antigens,
koff predicts the magnitude of TCR responsiveness in most studies (Bridgeman et al., 2012;
Govern et al., 2010; Huang et al., 2010; Kersh et al., 1998; Tian et al., 2007). Moreover,
mathematical modeling and experiments with TCR-induced IFNγ production showed that
biological potency correlated with an antigen's KD, but that maximal response was
determined by its koff (Dushek et al., 2011).
NIH-PA Author Manuscript
T cells expressing a monoclonal TCR together with a set of altered peptide ligands (APL)
are commonly used to address the issue of antigen discrimination by TCRs. OT-I is a murine
TCR recognizing MHCI (H2-Kb) loaded with OVA peptide (SIINFEKL) or OVA-derived
APLs. We previously showed that transgenic OT-I thymocytes discriminate between
positive and negative selecting APLs in a manner that was largely dependent on antigen
affinity and less dependent on a ligand concentration (Daniels et al., 2006). Two other
MHCI-restricted TCRs could similarly discriminate between negative and positive selecting
ligands. Threshold antigens were estimated to have on-cell KD ~ 6 μM and τ1/2 ~ 1 s
(Naeher et al., 2007; Palmer and Naeher, 2009). Moreover, negative, but not positive,
selectors provoked a sufficiently strong response in mature CD8 OT-I T cells to induce
autoimmunity in an experimental model of type I diabetes (King et al., 2012). These results
indicate that T lineage cells sense an intrinsic binding parameter of their antigens. One
plausible explanation includes the ability of a TCR to measure the duration of TCR-pMHC
interactions (antigen dwell-time), as suggested by a kinetic proofreading model of TCR
triggering (McKeithan, 1995). However, the mechanism used by the TCR to sense antigen
dwell-time is largely unknown.
In this study, we show that an antigen-engaged TCR scans multiple coreceptors to find one
that is coupled to Lck; this is the first and rate-limiting step in signal initiation. Based on
experimental data and mathematical modeling, we propose a mechanism of TCR signaling
(Lck come&stay/signal duration), where the kinetics of Lck delivery by coreceptors
underlies a kinetic proofreading process that establishes a dwell-time threshold for negative
Experimental Procedures
NIH-PA Author Manuscript
All adult mice were 6-12 weeks old and had a C57Bl/6 genetic background. OT-I Rag2−/−,
OT-I Rag2−/− β2m−/−, B3K508 Rag1−/− and B3K506 Rag1−/− mice were described
previously (Daniels et al., 2006; Huseby et al., 2005).
CD8.4 OT-I Rag2−/− β2m−/− strain was generated by crossing CD8.4 knock-in mouse
(Erman et al., 2006) with OT-I Rag2−/− β2m−/−. Mice were bred in our colony (University
Hospital Basel) in accordance with Cantonal and Federal laws of Switzerland.
Surface plasmon resonance
SPR equilibrium binding analysis was performed using a BIAcore T100™ equipped with a
CM5 sensor chip. SPR equilibrium analyses were carried out to determine the KD values for
OT-I:H2-Kb-APL interactions at 25°C. Ca. 300 response units of pMHC or TCR were
Cell. Author manuscript; available in PMC 2015 October 09.
Stepanek et al.
Page 4
coupled to the CM5 sensor chip surface. Analyte was injected at concentrations ranging
from 10 times above and 10 times below the estimated KD of the interaction.
NIH-PA Author Manuscript
Flow cytometry
Live cells were stained with relevant Abs on ice. For intracellular staining, cells were fixed
in 4% paraformaldehyde (15 min, RT), permeabilized by 90% methanol (30 min, on ice) and
stained with indicated antibodies at RT. Determination of surface molecule number was
performed using saturating concentrations of PE-conjugated Abs and calibration beads.
Calcium mobilization was measured using Indo-1 probe and Calcium response index was
calculated (see Extended Material and Methods). Flow cytometry immunoprecipitation (FCIP) was done as described previously (Schrum et al., 2007). Briefly, beads coated with antiCD4 or anti-CD8α Abs were used to pull down CD4 or CD8. Subsequently, PE-conjugated
anti-CD4, anti-CD8β, or anti-Lck Abs were used to quantify Lck/CD8 and Lck/CD4
coupling ratios.
NIH-PA Author Manuscript
Flow cytometry was carried out with a FACSCantoII (BD Bioscience). Cell sorting was
performed using an Influx sorter (BD Bioscience). Data were analyzed using FlowJo
software (TreeStar).
Determination of Lck phosphorylation status
Lck was immunoprecipitated from untreated, or PV or PP2 treated thymocytes and analyzed
by Western blotting. Signals from Abs recognizing phosphorylated and nonphosphorylated
Lck (Y394) were normalized to total Lck and the percentage of phosphorylated molecules
was calculated as previously described (Stepanek et al., 2011).
On-cell dwell time measurement
Qdot-pMHC monomers were added to T cells or thymocytes attached to polylysine-coated
borosilicate glass. Binding of Qdot-pMHC monomers was observed using single molecule
microscopy for 2.5 minutes and the duration of pMHC binding events was measured. The
number of persisting binding events was ploted vs. time and fitted with to a one phase
exponential decay function. See also Extended materials and methods.
NIH-PA Author Manuscript
FTOCs were performed as described (Hogquist et al., 1994). Briefly, thymic lobes were
excised from mice at a gestational age of day 15.5 and incubated in the presence of
particular peptide and, in case of OT-I Rag2−/− β2m−/− thymi, exogenous β2m (5 μg/ml).
After 7 days of culture, thymocytes were analyzed by flow cytometry.
Western blotting
Samples for Western blotting were heated in Laemmli sample buffer (2 min, 95 °C),
sonicated and analyzed by SDS-PAGE and Western blotting, using AlexaFluor790- or
AlexaFluor680-conjugated goat anti-mouse or anti-rabbit secondary antibodies (Jackson
ImmunoResearch). Specific Ab signals were quantified using an Odyssey Infrared Imaging
System (LI-COR Biosciences) and analysed by ImageJ software (NIH).
Cell. Author manuscript; available in PMC 2015 October 09.
Stepanek et al.
Page 5
Mathematical model
NIH-PA Author Manuscript
The models are composites of Markov chain model describing behaviour of surface
molecules, equations describing TCR occupancy, and equations describing initial catalytic
steps in TCR triggering. The models predict either a number of triggered TCRs within a
time interval or a number of occupied and triggered TCRs in any given time point.
Dwell time threshold for negative selection by pMHCI vs. pMHCII ligands
NIH-PA Author Manuscript
To identify positive and negative selecting ligands for MHCII-restricted TCRs, we analyzed
the development of I-Ab- restricted, peptide-specific B3K506 and B3K508 TCR transgenic
Rag1−/− thymocytes exposed to a variety of APLs with known affinities (Huseby et al.,
2006; Huseby et al., 2005). Using the frequency of CD4 single positive (SP) cells as an
indicator of negative selection in B3K508 fetal thymic organ cultures (FTOCs), we
identified ligands behaving as negative selectors (3K, P5R, and P2A), one partial negative
selector (threshold selector) (P-1A), and two ligands unable to negatively select (P3A, P-1K)
(Figure 1A-B). Interestingly, negative and positive selecting ligands also affected the
development of the CD8 SP compartment. Previous work showed that developing MHCIIrestricted thymocytes undergoing negative selection generate a population of CD8αα innatelike T cells, while the same cells undergoing positive selection paradoxically select a minor
population of CD8αβ SP cells (Yamagata et al., 2004). Similarly, in B3K508 FTOCs only
negative selecting ligands generated a population of CD8αα SP cells, while CD8αβ SP
thymocytes were present in FTOCs exposed to threshold and positive selecting ligands
(Figure S1A). Nevertheless, negative selection can be followed by the disappearance of the
CD4 SP population. We also identified negative and threshold selectors using FTOCs from
B3K506 Rag1−/− mice (Figure 1C). Threshold selectors for B3K508 and B3K506
thymocytes exhibited similar KD values (263 μM and 319 μM, respectively) (Huseby et al.,
2006), indicating that MHCII-restricted TCRs use a fixed affinity threshold (KD ~300 μM)
for negative selection.
NIH-PA Author Manuscript
To compare the affinities of threshold antigens mediating MHCI- and MHCII-restricted
negative selection, we determined the affinities of several antigens for MHC-I restricted OTI TCR using SPR (Figures 1D, S1B-D). The threshold selector for OT-I thymocytes (Kb-T4)
had a KD value of 444 μM (Figure 1D and (Daniels et al., 2006)). Taken together, results
from our FTOC and SPR experiments argue that MHCI- and MHCII-restricted TCRs use a
similar ligand affinity threshold to initiate negative selection. However, SPR affinity
measurements neglect the roles of CD4 and CD8 coreceptors that bind to pMHCII and
pMHCI antigens, respectively. Importantly, CD8, but not CD4, stabilizes the TCR-pMHC
complex and prolongs a ligand's dwell time on the cell surface (Huppa et al., 2010; Naeher
et al., 2007; Stone et al., 2009).
To directly determine the length of TCR-pMHC interactions in the presence of coreceptors,
we measured the dwell times of Qdot-labeled-pMHC monomers on their respective
peripheral T cells or double positive (DP) thymocytes by direct observation using single
molecule microscopy (Movie S1-2). The observed dwell times could be fitted to a one phase
Cell. Author manuscript; available in PMC 2015 October 09.
Stepanek et al.
Page 6
NIH-PA Author Manuscript
exponential decay curve (Figure 1E). For OT-I, an MHCI-restricted TCR, the τ1/2 of the
various APLs on mature T cells range from 1s to 16s and the τ1/2 correlates well with
antigenic potency (Daniels et al., 2006). The τ1/2 of the threshold selector, Kb-T4, is 1.3s.
Interestingly, the τ1/2s of strong negative selectors for the B3K506 (3K) and B3K508 (P1-A)
MHCII-restricted TCRs are 1.3s and 1.4s, respectively.
We could not directly measure the dwell time distribution of pMHCII threshold ligands, due
to the speed of image acquisition. To estimate the τ1/2 for an MHCII-restricted threshold
selector, we made some additional measurements.
We determined the τ1/2s of several ligands binding to their respective DP thymocytes
(Figures 1E, panels 2 and 8; Figure 1F). The τ1/2s on DP thymocytes and peripheral T cells
are in good agreement but the τ1/2s of a ligand binding to a thymocyte is ~1/3 shorter
(Figure 1F). As peripheral T cells have substantially more TCR, the calculated τ1/2s on these
cells may be slightly extended due to occasional rebinding to a second TCR. In any event,
the threshold τ1/2s for negative selection of MHCI restricted thymocytes is ~0.9s (Table S1).
NIH-PA Author Manuscript
We also observed that the on-cell koff and SPR KD are highly correlated, indicating that
differences in KD could be largely explained by differences in koff (Figure 1G). Therefore,
the extrapolated τ1/2 on thymocytes for pMHCII threshold antigens (KD ~ 300 μM) is ~ 0.2 s
(Table S1).
Thus, MHCI-restricted thymocytes use a longer dwell time threshold for negative selection
than MHCII-restricted thymocytes (0.9 vs. 0.2s). This raises the question, why MHCIIrestricted thymocytes initiate negative selection with a shorter dwell time.
Extent of coreceptor-Lck coupling determines the threshold for negative selection
CD4 binds Lck better than CD8 (Wiest et al., 1993), which could explain the shorter dwell
time threshold for MHCII- vs. MHCI-restricted thymocytes. We measured the CD4-Lck and
CD8-Lck coupling ratios in polyclonal pre-selection DP thymocytes (B6) using
immunoprecipitation followed by quantitative flow cytometric analysis (FC-IP). While only
0.6% of CD8αβ coreceptors bound Lck, 6.8% of CD4 molecules were Lck coupled (Figures
2A-B, S2A). The CD4-Lck coupling in B3K508 and B3K506 pre-selection DP thymocytes
was similar to that seen in polyclonal DPs from B6 mice (Figure S2B).
NIH-PA Author Manuscript
To further study the impact of coreceptor-Lck coupling on negative selection, we used
homozygous CD8.4 knock-in mice (Erman et al., 2006), which are unable to express
endogenous CD8α, and express instead only the chimeric CD8.4α chain, which consists of
the extracellular part of CD8α and a cytoplasmic CD4 tail, which binds Lck. In OT-I DP
thymocytes expressing CD8WT or CD8.4, coreceptor-Lck coupling is 1.1% and 5.8%,
respectively (Figure 2C-D). It was important to know the percentage of coreceptors coupled
with catalytically active Lck. For this reason we used pair of antibodies recognizing active
(pY394) and non-active (non-pY394) Lck, respectively (Nika et al., 2010). This analysis
indicated that the percentage of active Lck in the pre-selection DP thymocytes is 25-28%
(Figure 2E). Since the majority of Lck molecules are coreceptor-coupled (Van Laethem et
al., 2007), the fraction of CD4 and CD8 coreceptors loaded with catalytically active Lck in
Cell. Author manuscript; available in PMC 2015 October 09.
Stepanek et al.
Page 7
DP thymocytes is 1.8% and 0.16%, respectively. Thus, the majority of coreceptors in DP
thymocytes cannot initiate a TCR signal.
NIH-PA Author Manuscript
Expression of the CD8.4 coreceptor had no impact on the developmental arrest at the DP
stage or surface TCR levels in the OT-I DP thymocytes (Figure S2C). Moreover, CD8.4
does not significantly affect antigen binding since Kb-Q4R7 binds to CD8WT and CD8.4
OT-I DP thymocytes with a similar τ1/2 (Figure 1E, S2D). Despite these similarities, the
threshold for negative selection was strikingly reduced in CD8.4 thymocytes, converting
threshold and partial negative selectors (T4, Q4R7) into pure negative selectors and some
positive selectors (Q4H7, Q7) into threshold selectors (Figure 3A). The effect of increasing
CD8-Lck coupling cannot be mimicked by increasing the antigen concentration in CD8WT
OT-I FTOCs (Figure 3B). Ten-fold higher concentrations (20 μM) of the APLs in CD8WT
OT-I FTOCs do not generate the same degree of negative selection as 2 μM peptides used in
CD8.4 OT-I FTOCs. Therefore the increased Lck coupling to the CD8.4 coreceptor leads to
a shorter dwell time threshold for negative selection that is largely concentration insensitive.
The impact of Lck-coupling on the development of thymocytes is summarized in Table S1.
Increased CD8-Lck coupling enhances proximal signaling and cellular responses
NIH-PA Author Manuscript
We further focused on the role of Lck coupling in the initiation of TCR signaling by
analyzing the response of CD8WT and CD8.4 OT-I DP thymocytes to stimulation with
various Kb-peptide tetramers (strong negative selector, Kb-OVA; just above threshold
selector, Kb-Q4R7; positive selector, Kb-Q4H7). In CD8WT OT-I DP thymocytes, ligandinduced phosphorylation was modest for TCRζ and ZAP70 but down-stream signaling
proteins LAT, SLP76, VAV, and Erk1 exhibited more extensive phosphorylation upon
stimulation (Figure 4 A-G, S3A-C). The data indicate an amplification step between ZAP70
activation and LAT phosphorylation. Importantly, CD8.4 thymocytes exhibited enhanced
phosphorylation of TCRζ, ZAP70, LAT, SLP76, VAV, and Erk1 following TCR stimulation
(Figure 4 A-G, S3A-C). The CD8.4 coreceptor increased overall TCR proximal signaling by
1.4, 3.7, and 5.8 following stimulation with Kb-OVA, -Q4R7, and -Q4H7, respectively
(Figure S3D). Enhanced coreceptor-Lck coupling augments the proximal TCR signaling
especially for lower affinity ligands.
NIH-PA Author Manuscript
Tetramer stimulated CD8.4 thymocytes exhibited substantially increased calcium influx
compared to CD8WT thymocytes (Figure 4H). In fact, CD8.4 thymocytes stimulated with
below threshold Kb-Q4H7 tetramers exhibited a comparable level of calcium signaling as
CD8WT thymocytes stimulated with high affinity Kb-OVA tetramers, indicating a
substantial shift in the signaling threshold induced by the chimeric co-receptor (compare
purple and red curves in Figure 4H, upper left panel).
To address the role of CD8-Lck coupling in response to a broader range of APLs, CD8WT
and CD8.4 OTI DP thymocytes were stimulated with antigen presenting cells (APCs) loaded
with various peptide variants. We monitored induced expression of CD69, an activation
marker that is upregulated both upon positive and negative selection (Figure S4). While
CD8WT and CD8.4 thymocytes responded similarly to the strongest peptide, OVA, CD8.4
thymocytes were more sensitive to weaker antigens (Figure 5A-C). The impact of enhanced
Lck delivery mediated by the CD8.4 coreceptor inversely correlated with potency of the
Cell. Author manuscript; available in PMC 2015 October 09.
Stepanek et al.
Page 8
ligand (Figure 5D), consistent with the effects of CD8.4 on proximal signaling (Figure
NIH-PA Author Manuscript
Frequency of coreceptor-Lck coupling determines the kinetics of Lck delivery to the TCR
NIH-PA Author Manuscript
The overall experimental data implied that Lck delivery is a limiting factor in signal
initiation, and potentially sets the dwell time threshold to initiate negative selection. To
better understand why coreceptor-Lck coupling is so important in this process, we generated
a mathematical model that calculates the probability of recruiting an Lck-coupled coreceptor
to an established TCR:pMHC pair as a function of time, when the CD4, CD8, or CD8.4
coreceptor is involved. The model is based on a Markov chain that describes the behavior of
coreceptor and TCR-pMHC in the plasma membrane. A TCR-pMHC can form a pair (close
proximity) or a complex (binding) with an empty or Lck-coupled coreceptor (Figure 6A).
Because Lck-coupled coreceptors are relatively rare, a TCR-pMHC usually has to scan a
large number of empty coreceptors before encountering an Lck-coupled one. The average
time to recruit an Lck loaded coreceptor to the TCR-pMHC is determined by the rate of
coreceptor:TCR-pMHC complex formation, the duration of empty coreceptor:TCR-MHC
interaction, and importantly the percentage of Lck-coupled coreceptors (see Extended
Materials and Methods). The parameters used to construct the model were based on our
measurements or published data (Table S2, Figure S5A-C). The model shows that Lck
delivery is very fast in case of TCR-pMHCII interactions (CD4), but substantially slower
upon TCR-pMHCI engagement (CD8). CD8.4 shows faster Lck recruitment than CD8WT
(Figure 6B). When only active (pY394) Lck delivery is considered, the delivery of the Lck
is delayed but the hierarchy of the coreceptors is maintained (dashed line in Figure 6B).
Importantly, the model predicts that the probability of Lck recruitment is highly dependent
on the extent of coreceptor-Lck coupling (Figure 6C), underscoring its importance for
setting the dwell time threshold. Other parameters (number of CD8 molecules, diffusion
coefficient of membrane anchored receptors, coreceptor-TCR interaction kinetics, and lattice
spacing) play less important roles (Figure 6D).
NIH-PA Author Manuscript
To provide experimental evidence that a single TCR-pMHC complex serially engages large
number of co-receptors, we carried out the following experiment. Qdot-labeled Kb-OVA or
Kb-Q4R7 monomers were bound to OT-I DP thymocytes allowing the co-engagement of the
TCR and CD8 with the labeled antigen. After reaching equilibrium binding, the cells were
diluted in presence or absence of an anti-CD8β mAb. Blocking the unengaged CD8
molecules accelerated the off-rate of pMHC dissociation (Figure S5E-F). This is consistent
with the idea that during the life-time of a pMHC-engaged TCR, there is significant turnover
of CD8 molecules, which only transiently bind to the pMHC. This co-receptor exchange
would allow the engaged TCR to eventually find an Lck coupled co-receptor if the pMHC
engagement is sufficiently long.
Lck delivery is the initial step during kinetic proofreading
The Markov chain model explained the importance of the proportion of Lck-coupled
coreceptors and coreceptor exchange in determining the speed of the signal initiation. We
subsequently generated four different models that extended the Markov chain model by
combining Lck delivery together with its kinase activity. Comparison of the models’
Cell. Author manuscript; available in PMC 2015 October 09.
Stepanek et al.
Page 9
predictions with the experimental data should indicate the actual molecular mechanism that
initiates TCR triggering and sets the antigen dwell time threshold for negative selection.
NIH-PA Author Manuscript
For all four models, we assumed that the recruited Lck has to minimally phosphorylate 4
tyrosines (2 ITAM tyrosines and Y315 and Y319 in ZAP70) to obtain ZAP70 activation and
signal propagation. The actual number of tyrosine phosphorylations is likely higher because
phosphatase activity can counteract the Lck activity and Lck might phosphorylate tyrosines
in separate ITAMs. Thus, we set 5 phosphorylations by an Lck molecule as a criterion for
TCR triggering to make it more than the minimal number.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Our preferred model, ‘Lck come&stay/signal duration’ combines the Lck recruitment
calculated from the Markov chain model with the classical TCR kinetic proofreading model
(McKeithan, 1995). In this model, we assumed the following scenario. A newly formed
TCR-pMHC pair eventually interacts with a coreceptor coupled to active Lck (pLck). The
coreceptor-pLck remains attached to the TCR-pMHC complex due to the interactions of
Lck's SH2 domains with partially phosphorylated TCR (and/or ZAP70) and stabilization via
the pMHC:coreceptor interaction (Jiang et al., 2011; Straus et al., 1996). Lck triggers TCR
signaling by phosphorylating ITAM tyrosines and a subsequently recruited ZAP70. Active
ZAP70 continuously generates down-stream signals by phosphorylating LAT, SPL76 and
other signaling molecules. When antigen disengages from the TCR, coreceptor-Lck is
released as well, leading to a massive decrease in kinase activity. This shifts the kinase/
phosphatase equilibrium towards phosphatase activity and sets the TCR and ZAP-70 back
into a less phosphorylated state. ‘Lck come&stay/signal duration’ model assumes that the
TCR-pMHC interaction has to last long enough to recruit and activate ZAP70. Once the
ZAP70 is activated, further antigen occupancy of the already triggered TCR determines the
strength of the TCR signal. The overall TCR signal induced in a thymocyte is determined by
the number of triggered TCRs and the amount of time they remain occupied by pMHC
subsequent to TCR triggering (Extended methods and Table S3). The model calculates total
TCR signal as a function of number of antigen molecules available in the thymocyte/APC
interface and their τ1/2 (Figure 7A, S6A). In contrast to a pure TCR occupancy model
(Figure 7B) that predicts little discrimination between ligands with different dwell times, the
‘Lck come&stay/signal duration’ model displays the kinetic proofreading principle. Our data
clearly show that antigens with τ1/2 ~ 10s (Kb-OVA) can negatively select at low
concentrations; antigens with τ1/2 ~1-2s (Kb-T4 and Kb-Q4R7) negatively select at higher
concentrations; and antigens with τ1/2 ≤ 0.6s (Kb-Q4H7) are incapable of negative selection
even when present at very high concentrations in the synapse (i.e. 250 cognate pMHC,
which is roughly 5-10% of the pMHCI pool at the interface). The model assumes that as few
as 2-4 triggered and occupied TCRs can initiate negative selection, which has previous
experimental support (Ebert et al., 2008; Peterson et al., 1999).
The model predicts a shift of the dwell time threshold for negative selection in CD8.4 cells
(Figure 7A, C), which in good agreement with the experimental data (Figure 1, 3, Table S1).
Moreover, this model predicts that the shorter the τ1/2 of a TCR ligand is, the more
pronounced is the difference between CD8.4 and CD8WT signaling (Figure 7D). This is
essentially in agreement with the experimental data (Figure 4-5). The ‘Lck come&stay/
signal duration’ model predicts the dwell time thresholds used by polyclonal pMHCI- and
Cell. Author manuscript; available in PMC 2015 October 09.
Stepanek et al.
Page 10
NIH-PA Author Manuscript
pMHCII-restricted thymocytes (~1s for MHCI and ~0.2 for MHCII antigens) (Figure 7E).
The model also reveals that antigens with τ1/2 at or just above the dwell time threshold must
be presented at a relatively high copy number to induce negative selection (e.g. 50
molecules of pMHCI antigens with τ1/2 ~ 2 s). If such peri-threshold ligands are presented at
low numbers in the thymus, but at higher levels in peripheral tissues, the generation of
central self-tolerance may be defective.
The serial triggering (productive hit) model assumes that once a TCR is triggered, continued
pMHC binding does not further increase the TCR signal (Dushek et al., 2011; Valitutti,
2012). We also constructed a combined ‘Lck come&stay/serial triggering’ model which
assumes that the total TCR signal is determined by the number of triggered TCRs. However,
the ‘Lck come&stay/serial triggering’ model does not predict the experimentally observed
hierarchy of the ligands, arguing that serial triggering cannot explain the negative selection
threshold (Figure S6B).
NIH-PA Author Manuscript
We also generated two additional models requiring multiple visits of Lck to trigger the TCR
(Figure S6C). In these models, we assumed that Lck is not stabilized at the TCR-pMHC and
phosphorylates only one tyrosine upon a single interaction between a coreceptor-Lck
molecule and the TCR-pMHC. Thus, coreceptor-Lck has to be repeatedly recruited to the
TCR-pMHC to trigger the TCR. These models predict very strong discrimination between
ligands based on their τ1/2, but displays very low sensitivity to ligands, just over the dwelltime threshold; this is not compatible with our experimental data. Given the poor predictive
power of these last three models, the ‘Lck come&stay/signal duration’ model clearly
provides the best explanation of our experimental data.
NIH-PA Author Manuscript
Here we describe a mechanism, which allows thymocytes to discriminate antigens of
differing median dwell times and establish an antigen dwell time threshold for negative
selection. During the time that a TCR binds a pMHC antigen, the TCR-pMHC complex
scans up to several hundreds of coreceptors to find one, which carries catalytically active
Lck; subsequently this recruited kinase phosphorylates ITAMs and activating tyrosines on
ZAP70. Finding an Lck-loaded coreceptor represents the first and rate-limiting step in signal
initiation, because only a minority of co-receptor molecules are actually coupled to active
Lck. MHCI-restricted TCRs require antigens with a τ1/2 > 0.9s to induce negative selection,
while the negative selection threshold for MHCII-restricted TCRs is τ1/2 > 0.2s. A higher
frequency of CD4 molecules are coupled to catalytically active Lck compared to CD8 (2%
vs 0.2%); for this reason, MHCII-restricted receptors have a shorter dwell time threshold.
Modeling studies revealed that the frequency of Lck-coupled coreceptors is a critical
parameter in establishing the dwell time threshold for negative selection. The decrease in the
dwell time threshold for negative selection seen in CD8.4 OT-I mice is a consequence of
CD8.4's increased coupling to Lck, resulting in a faster arrival of a coreceptor-pLck complex
to an antigen-occupied TCR. CD8.4 mediates an enhanced activation of proximal TCR
signaling pathways and, similar to CD4, allows negative selection by antigens with short
dwell times. It's not clear whether having a shorter dwell time threshold for the deletion of
Cell. Author manuscript; available in PMC 2015 October 09.
Stepanek et al.
Page 11
NIH-PA Author Manuscript
MHCII-restricted thymocytes provides an intrinsic advantage or whether this is simply a
consequence of the differences in Lck coupling exhibited by CD4 and CD8. Nevertheless,
these differences illustrate the mechanism, where the dwell time threshold is set by the
median time required for the antigen engaged TCR to find an Lck-loaded co-receptor.
NIH-PA Author Manuscript
To directly estimate the dwell-time threshold, we attached monomeric pMHCs to Qdots and
observed the binding of monomeric pMHC to antigen specific T cells by single molecule
microscopy. The τ1/2 for the OT-I threshold ligand (Kb-T4) was estimated to 0.9s on
thymocytes and 1.3s on peripheral T cells, which corresponds well to the dwell time
threshold for H2-Kd-restricted T1 TCR (0.8-1.5s), estimated by an on-cell photoaffinity
labeling technique (Naeher et al., 2007; Palmer and Naeher, 2009). How well does τ1/2
measured with soluble ligand (3D) correspond to antigen binding at the thymocyte/APC
interface? Although the τ1/2 was originally viewed as a biophysical parameter describing
ligand-receptor interactions independent of their spatial context (Dustin et al., 2001), several
studies report that koff is accelerated in a planar configuration (2D) of T cell-APC contacts
(Huang et al., 2010; Huppa et al., 2010). However, other studies showed that pMHC τ1/2
measured using 2D or 3D techniques are similar, consistent with the original view
(O'Donoghue et al., 2013; Robert et al., 2012). Recently, pMHCI/TCR dwell times were
analyzed under tensile force in a 2D setting (Liu et al., 2014). Interestingly, TCR
interactions with potent ligands are stabilized when a moderate force (5-10 pN) is applied
due to the formation of catch bonds. The authors suggest that a tensile force generated at the
thymocyte/APC interface contribute to TCR specificity. However, their results show that
both positive and negative selectors can form catch bonds, so the formation of a catch bond
per se does not explain the negative selection threshold. Nevertheless, the catch bond
phenomenon can be easily integrated with our ‘Lck come&stay/signal duration model’ that
provides the signaling mechanism underlying TCR triggering and discrimination between
negative and positive ligands based on their dwell times. Interestingly, τ1/2 measurements in
their system corresponded well with our on-cell measurements. We determined that the τ1/2
of OT-I:Kb-OVA is ~10s, while Liu et al. measured the τ1/2 for the same antigen as 0.8s
under 10 pN; the ~10 fold difference is expected because of the absence of CD8 engagement
in their system (Naeher et al., 2007). However, it remains unclear, why an application of a
moderate tensile force is required in the 2D setting, but not in our on-cell 3D assay, to
measure biologically relevant dwell times of TCR-antigen interactions.
NIH-PA Author Manuscript
We generated four mathematical models describing TCR signaling response in thymocytes,
using a combination of our measurements and published data as input parameters. The kon
(Huppa et al., 2010) was sufficiently high that it hardly limited the formation of TCR-pMHC
bonds in the thymocyte:APC contact area. Ligand concentration and τ1/2 had the largest
impact on the quantity of TCR signals generated. To find the most relevant model, we
compared the outcome of the models with our experimental data on TCR signaling in
CD8WT and CD8.4 OT-I DP thymocytes. The ‘Lck come&stay/Signal duration’ model,
which best explains our experimental data, is based on the principle of kinetic proofreading
(McKeithan, 1995), where Lck recruitment is the most proximal limiting step. The model
assumes that transient CD4-MHCII or CD8-MHCI interactions allow the TCR-pMHC pair
to scan multiple coreceptors via coreceptor exchange before finding a coreceptor carrying
Cell. Author manuscript; available in PMC 2015 October 09.
Stepanek et al.
Page 12
NIH-PA Author Manuscript
catalytically active Lck. Nika et al. demonstrated that 38% of Lck molecules are
catalytically active in human CD4+ T cells and moreover Lck does not undergo further
activation upon TCR stimulation (Nika et al., 2010). Murine preselection DP thymocytes
have a slightly lower percentage of active Lck (25-28%), implying that only ~0.16% of CD8
and ~1.8% of CD4 molecules are capable of initiating a TCR signal. Following the
recruitment of catalytically active Lck, the kinase phosphorylates ITAM tyrosines in the
CD3 complex and subsequently phosphorylates a recruited ZAP70 kinase. Coreceptor-Lck
binding to TCR-pMHC is potentially stabilized via an interaction of Lck's SH2 domain with
a phosphotyrosine in an ITAM or ZAP70 (Jiang et al., 2011; Straus et al., 1996). If the TCRpMHC interaction lasts long enough to enable Lck-mediated phosphorylation of ZAP70,
then ZAP70 generates downstream signals by phosphorylating LAT and SLP76 for the
duration of the TCR-pMHC interaction. The longer the duration of antigen binding, the
more downstream mediators are generated. Signaling is terminated when the pMHC and
subsequently, the coreceptor-Lck disengage from TCR. A thymocyte surveys a single APC
for several minutes (Ebert et al., 2008; Melichar et al., 2013); over this time span, the
decision to initiate negative selection must be made.
NIH-PA Author Manuscript
We propose that positive and negative selecting ligands induce quantitatively different
responses at the level of TCR activation (i.e. number of TCRs kept triggered) that is
transformed to qualitatively different events in down-stream signaling pathways. Our model
is in line with the observation that positive and negative selecting ligands induce distinct
patterns of Erk activation (Daniels et al., 2006). Previous mathematical modeling suggested
that only negative selecting ligands are able to induce extensive LAT phosphorylation and
trigger a feed-forward loop resulting in the activation of SOS and activation of pErk at the
plasma membrane (Prasad et al., 2009).
NIH-PA Author Manuscript
The TCR has the unusual property of being able to recognize antigens with a high degree of
sensitivity and an ability to discriminate between closely related structural variants. As few
as 2-4 strong ligands within the thymocyte-APC contact area are able to trigger a strong
TCR response and induce negative selection (Ebert et al., 2008; Peterson et al., 1999). This
sensitivity has been explained by the ability of one pMHC ligand to trigger several TCRs
and amplify the signal. However, evidence for serial triggering is so far indirect (Valitutti,
2012). Moreover, a strict serial triggering model does not explain our experimental data,
since it predicts that threshold ligands (Kb-T4) will engage and trigger similar number of
TCRs as long dwelling ligands (Kb-OVA). Furthermore, serial triggering could not be
detected during direct observation of TCR-pMHC interactions in situ (O'Donoghue et al.,
2013). Along this line, we propose that very few strong ligands are able to keep a few TCRs
continually engaged and triggered; this amplifies the TCR signal by continuous generation
of intracellular signaling intermediates. Higher numbers of less potent, but above threshold
ligands are required to induce the same effect, while below threshold ligands are essentially
unable to trigger negative selection.
The two models, which employ the Lck multiple visit assumption, are relatively insensitive.
These models predict that only antigens with very long dwell times or present at high
concentrations can initiate a TCR signal. For this reason, we do not favor these models. We
also observed that the serial triggering strategy poorly discriminates between antigens with
Cell. Author manuscript; available in PMC 2015 October 09.
Stepanek et al.
Page 13
NIH-PA Author Manuscript
differing median dwell times. On the other hand, the signal duration mechanism allows the
amount of TCR generated signal to reflect the actual dwell time. Thus, the ‘Lck come and
stay/signal duration’ mechanism allows for antigen recognition, which is both highly
sensitive and discriminatory. ‘Lck come and stay/signal duration’ model uncovers molecular
events constituting for a TCR kinetic proofreading mechanism. Highly specific recognition
of negative selecting ligands by a developing thymocyte is an analogy to a kinetic
proofreading that provides specificity in DNA, protein, and amino acid-tRNA synthesis
(Hopfield, 1974). These diverse molecular processes have a built in time delay between a
substrate binding the enzyme and its conversion into a product molecule. A substrate
molecule (nucleotide, amino acid etc.) that binds the respective enzyme for a longer time is
more likely to be the biologically ‘correct’ molecular species. The TCR and coreceptor
operate under a similar principle; a long lasting antigen binding event is more likely to be
converted into a (negative selection) signaling event.
NIH-PA Author Manuscript
Our experimental results and mathematical models explain, how the TCR actually measures
antigen affinity to initiate a negative selection signal. The kinetics of Lck delivery by
coreceptors plays a crucial role in setting the antigen dwell time threshold for negative
selection. The kinetic proofreading mechanism suggested by our model implies that
collisions between hundreds of coreceptor molecules and a small number of antigen bound
TCRs allow the developing thymocyte to sense the antigen's median dwell time and initiate
a critical cell fate decision (negative selection).
Supplementary Material
Refer to Web version on PubMed Central for supplementary material.
We thank U. Schneider for animal husbandry, and E. Traunecker and T. Krebs for cell sorting. This study was
funded by grants to EP (310030-149972/1 and Synergia [SNF], Sybilla [EU FP7], and TerraIncognita [ERC]), AKC
(1-P01-AI071195 [NIH]), ESH (R01-DK095077 [NIH]), and AKS (BB/H001085/1[BBSRC]). DKC is a Wellcome
Trust Career Development Fellow. AKS is a Wellcome Trust Senior Investigator.
NIH-PA Author Manuscript
Artyomov MN, Lis M, Devadas S, Davis MM, Chakraborty AK. CD4 and CD8 binding to MHC
molecules primarily acts to enhance Lck delivery. Proc Natl Acad Sci U S A. 2010; 107:16916–
16921. [PubMed: 20837541]
Bridgeman JS, Sewell AK, Miles JJ, Price DA, Cole DK. Structural and biophysical determinants of
alphabeta T-cell antigen recognition. Immunology. 2012; 135:9–18. [PubMed: 22044041]
Daniels MA, Teixeiro E, Gill J, Hausmann B, Roubaty D, Holmberg K, Werlen G, Hollander GA,
Gascoigne NR, Palmer E. Thymic selection threshold defined by compartmentalization of Ras/
MAPK signalling. Nature. 2006; 444:724–729. [PubMed: 17086201]
Dushek O, Aleksic M, Wheeler RJ, Zhang H, Cordoba SP, Peng YC, Chen JL, Cerundolo V, Dong T,
Coombs D, et al. Antigen potency and maximal efficacy reveal a mechanism of efficient T cell
activation. Sci Signal. 2011; 4:ra39. [PubMed: 21653229]
Dustin ML, Bromley SK, Davis MM, Zhu C. Identification of self through two-dimensional chemistry
and synapses. Annu Rev Cell Dev Biol. 2001; 17:133–157. [PubMed: 11687486]
Cell. Author manuscript; available in PMC 2015 October 09.
Stepanek et al.
Page 14
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Ebert PJ, Ehrlich LI, Davis MM. Low ligand requirement for deletion and lack of synapses in positive
selection enforce the gauntlet of thymic T cell maturation. Immunity. 2008; 29:734–745. [PubMed:
Erman B, Alag AS, Dahle O, van Laethem F, Sarafova SD, Guinter TI, Sharrow SO, Grinberg A, Love
PE, Singer A. Coreceptor signal strength regulates positive selection but does not determine
CD4/CD8 lineage choice in a physiologic in vivo model. J Immunol. 2006; 177:6613–6625.
[PubMed: 17082573]
Govern CC, Paczosa MK, Chakraborty AK, Huseby ES. Fast on-rates allow short dwell time ligands
to activate T cells. Proc Natl Acad Sci U S A. 2010; 107:8724–8729. [PubMed: 20421471]
Hogquist KA, Jameson SC, Heath WR, Howard JL, Bevan MJ, Carbone FR. T cell receptor antagonist
peptides induce positive selection. Cell. 1994; 76:17–27. [PubMed: 8287475]
Hopfield JJ. Kinetic proofreading: a new mechanism for reducing errors in biosynthetic processes
requiring high specificity. Proc Natl Acad Sci U S A. 1974; 71:4135–4139. [PubMed: 4530290]
Huang J, Zarnitsyna VI, Liu B, Edwards LJ, Jiang N, Evavold BD, Zhu C. The kinetics of twodimensional TCR and pMHC interactions determine T-cell responsiveness. Nature. 2010;
464:932–936. [PubMed: 20357766]
Huppa JB, Axmann M, Mortelmaier MA, Lillemeier BF, Newell EW, Brameshuber M, Klein LO,
Schutz GJ, Davis MM. TCR-peptide-MHC interactions in situ show accelerated kinetics and
increased affinity. Nature. 2010; 463:963–967. [PubMed: 20164930]
Huseby ES, Crawford F, White J, Marrack P, Kappler JW. Interface-disrupting amino acids establish
specificity between T cell receptors and complexes of major histocompatibility complex and
peptide. Nat Immunol. 2006; 7:1191–1199. [PubMed: 17041605]
Huseby ES, White J, Crawford F, Vass T, Becker D, Pinilla C, Marrack P, Kappler JW. How the T cell
repertoire becomes peptide and MHC specific. Cell. 2005; 122:247–260. [PubMed: 16051149]
Jiang N, Huang J, Edwards LJ, Liu B, Zhang Y, Beal CD, Evavold BD, Zhu C. Two-stage cooperative
T cell receptor-peptide major histocompatibility complex-CD8 trimolecular interactions amplify
antigen discrimination. Immunity. 2011; 34:13–23. [PubMed: 21256056]
Kerry SE, Buslepp J, Cramer LA, Maile R, Hensley LL, Nielsen AI, Kavathas P, Vilen BJ, Collins EJ,
Frelinger JA. Interplay between TCR affinity and necessity of coreceptor ligation: high-affinity
peptide-MHC/TCR interaction overcomes lack of CD8 engagement. J Immunol. 2003; 171:4493–
4503. [PubMed: 14568922]
Kersh GJ, Kersh EN, Fremont DH, Allen PM. High- and low-potency ligands with similar affinities
for the TCR: the importance of kinetics in TCR signaling. Immunity. 1998; 9:817–826. [PubMed:
King CG, Koehli S, Hausmann B, Schmaler M, Zehn D, Palmer E. T cell affinity regulates asymmetric
division, effector cell differentiation, and tissue pathology. Immunity. 2012; 37:709–720.
[PubMed: 23084359]
McKeithan TW. Kinetic proofreading in T-cell receptor signal transduction. Proc Natl Acad Sci U S
A. 1995; 92:5042–5046. [PubMed: 7761445]
Melichar HJ, Ross JO, Herzmark P, Hogquist KA, Robey EA. Distinct temporal patterns of T cell
receptor signaling during positive versus negative selection in situ. Sci Signal. 2013; 6:ra92.
[PubMed: 24129702]
Naeher D, Daniels MA, Hausmann B, Guillaume P, Luescher I, Palmer E. A constant affinity
threshold for T cell tolerance. J Exp Med. 2007; 204:2553–2559. [PubMed: 17938233]
Nika K, Soldani C, Salek M, Paster W, Gray A, Etzensperger R, Fugger L, Polzella P, Cerundolo V,
Dushek O, et al. Constitutively active Lck kinase in T cells drives antigen receptor signal
transduction. Immunity. 2010; 32:766–777. [PubMed: 20541955]
O'Donoghue GP, Pielak RM, Smoligovets AA, Lin JJ, Groves JT. Direct single molecule measurement
of TCR triggering by agonist pMHC in living primary T cells. Elife. 2013; 2:e00778. [PubMed:
Palmer E, Naeher D. Affinity threshold for thymic selection through a T-cell receptor-co-receptor
zipper. Nat Rev Immunol. 2009; 9:207–213. [PubMed: 19151748]
Cell. Author manuscript; available in PMC 2015 October 09.
Stepanek et al.
Page 15
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Peterson DA, DiPaolo RJ, Kanagawa O, Unanue ER. Cutting edge: negative selection of immature
thymocytes by a few peptide-MHC complexes: differential sensitivity of immature and mature T
cells. J Immunol. 1999; 162:3117–3120. [PubMed: 10092759]
Prasad A, Zikherman J, Das J, Roose JP, Weiss A, Chakraborty AK. Origin of the sharp boundary that
discriminates positive and negative selection of thymocytes. Proc Natl Acad Sci U S A. 2009;
106:528–533. [PubMed: 19098101]
Robert P, Aleksic M, Dushek O, Cerundolo V, Bongrand P, van der Merwe PA. Kinetics and
mechanics of two-dimensional interactions between T cell receptors and different activating
ligands. Biophys J. 2012; 102:248–257. [PubMed: 22339861]
Schrum AG, Gil D, Dopfer EP, Wiest DL, Turka LA, Schamel WW, Palmer E. High-sensitivity
detection and quantitative analysis of native protein-protein interactions and multiprotein
complexes by flow cytometry. Sci STKE. 2007; 2007:l2.
Smith-Garvin JE, Koretzky GA, Jordan MS. T cell activation. Annu Rev Immunol. 2009; 27:591–619.
[PubMed: 19132916]
Stepanek O, Kalina T, Draber P, Skopcova T, Svojgr K, Angelisova P, Horejsi V, Weiss A, Brdicka T.
Regulation of Src family kinases involved in T cell receptor signaling by protein-tyrosine
phosphatase CD148. J Biol Chem. 2011; 286:22101–22112. [PubMed: 21543337]
Stone JD, Chervin AS, Kranz DM. T-cell receptor binding affinities and kinetics: impact on T-cell
activity and specificity. Immunology. 2009; 126:165–176. [PubMed: 19125887]
Straus DB, Chan AC, Patai B, Weiss A. SH2 domain function is essential for the role of the Lck
tyrosine kinase in T cell receptor signal transduction. J Biol Chem. 1996; 271:9976–9981.
[PubMed: 8626636]
Straus DB, Weiss A. Genetic evidence for the involvement of the lck tyrosine kinase in signal
transduction through the T cell antigen receptor. Cell. 1992; 70:585–593. [PubMed: 1505025]
Straus DB, Weiss A. The CD3 chains of the T cell antigen receptor associate with the ZAP-70 tyrosine
kinase and are tyrosine phosphorylated after receptor stimulation. J Exp Med. 1993; 178:1523–
1530. [PubMed: 7693848]
Tian S, Maile R, Collins EJ, Frelinger JA. CD8+ T cell activation is governed by TCR-peptide/MHC
affinity, not dissociation rate. J Immunol. 2007; 179:2952–2960. [PubMed: 17709510]
Valitutti S. The Serial Engagement Model 17 Years After: From TCR Triggering to Immunotherapy.
Front Immunol. 2012; 3:272. [PubMed: 22973273]
van der Merwe PA, Dushek O. Mechanisms for T cell receptor triggering. Nat Rev Immunol. 2011;
11:47–55. [PubMed: 21127503]
Van Laethem F, Sarafova SD, Park JH, Tai X, Pobezinsky L, Guinter TI, Adoro S, Adams A, Sharrow
SO, Feigenbaum L, et al. Deletion of CD4 and CD8 coreceptors permits generation of alphabetaT
cells that recognize antigens independently of the MHC. Immunity. 2007; 27:735–750. [PubMed:
Van Laethem F, Tikhonova AN, Pobezinsky LA, Tai X, Kimura MY, Le Saout C, Guinter TI, Adams
A, Sharrow SO, Bernhardt G, et al. Lck Availability during Thymic Selection Determines the
Recognition Specificity of the T Cell Repertoire. Cell. 2013; 154:1326–1341. [PubMed:
Veillette A, Bookman MA, Horak EM, Bolen JB. The CD4 and CD8 T cell surface antigens are
associated with the internal membrane tyrosine-protein kinase p56lck. Cell. 1988; 55:301–308.
[PubMed: 3262426]
Vidal K, Daniel C, Hill M, Littman DR, Allen PM. Differential requirements for CD4 in TCR-ligand
interactions. J Immunol. 1999; 163:4811–4818. [PubMed: 10528181]
Wiest DL, Yuan L, Jefferson J, Benveniste P, Tsokos M, Klausner RD, Glimcher LH, Samelson LE,
Singer A. Regulation of T cell receptor expression in immature CD4+CD8+ thymocytes by p56lck
tyrosine kinase: basis for differential signaling by CD4 and CD8 in immature thymocytes
expressing both coreceptor molecules. J Exp Med. 1993; 178:1701–1712. [PubMed: 8228817]
Williams CB, Engle DL, Kersh GJ, Michael White J, Allen PM. A kinetic threshold between negative
and positive selection based on the longevity of the T cell receptor-ligand complex. J Exp Med.
1999; 189:1531–1544. [PubMed: 10330432]
Cell. Author manuscript; available in PMC 2015 October 09.
Stepanek et al.
Page 16
NIH-PA Author Manuscript
Yamagata T, Mathis D, Benoist C. Self-reactivity in thymic double-positive cells commits cells to a
CD8 alpha alpha lineage with characteristics of innate immune cells. Nat Immunol. 2004; 5:597–
605. [PubMed: 15133507]
Yin L, Dai S, Clayton G, Gao W, Wang Y, Kappler J, Marrack P. Recognition of self and altered self
by T cells in autoimmunity and allergy. Protein Cell. 2013; 4:8–16. [PubMed: 23307779]
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Cell. Author manuscript; available in PMC 2015 October 09.
Stepanek et al.
Page 17
NIH-PA Author Manuscript
TCR scans multiple co-receptors to initiate a TCR signal and negative selection.
Frequency of Lck loaded co-receptors sets antigen dwell time threshold.
Shorter antigen dwell time threshold for pMHCII compared to pMHCI selfantigens.
‘Lck come & stay/signal duration’ model explains how the TCR reads antigen
dwell time.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Cell. Author manuscript; available in PMC 2015 October 09.
Stepanek et al.
Page 18
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Figure 1. Thresholds for negative selection by pMHCII and pMHCI ligands show similar SPR
affinities but different on-cell dwell times
Fetal thymi from B3K508 Rag1−/− (A-B), B3K506 Rag1−/− (C), and OTI Rag2−/− β2m−/−
mice (D) were cultured with different APLs (20 μM) for 7 days and stained for CD4 and
CD8. (A) Effects of 3K, P1-A, and P3A on the B3K508 Rag1−/− thymocyte development.
Each panel is a representative plot from 2 thymi. (B-D) Percentage of CD4 or CD8β single
positive cells versus KD of APLs. Mean ± range, n=2-5. Square symbol shows percentage of
single positive cells without peptide (mean ± SEM, n = 5-11). (E) Distribution of dwell
Cell. Author manuscript; available in PMC 2015 October 09.
Stepanek et al.
Page 19
NIH-PA Author Manuscript
times of pMHCI and pMHCII ligands on TCR transgenic CD4 or CD8 peripheral T cells, or
preselection DP thymocytes. Data were fitted using one phase exponential decay curve. (F)
τ1/2 on preselection thymocytes vs. τ1/2 on peripheral cells were plotted and fitted using
linear regression with a fixed [0;0] point. (G) koff, calculated from on-cell τ1/2 on peripheral
T cells vs. KD determined by SPR was plotted for pMHCI and pMHCII ligands and fitted
with a linear regression with fixed [0;0] point. See also Figure S1 and Table S1.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Cell. Author manuscript; available in PMC 2015 October 09.
Stepanek et al.
Page 20
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Figure 2. Quantitative determination of Lck coupling to CD4, CD8, and CD8.4 coreceptors
Cell lysates were incubated with beads coated with antibodies to CD4, CD8β, or isotype
controls. Beads were probed with PE-conjugated antibodies to Lck, CD8α, or CD4 and
analyzed by flow cytometry. (A-B) Sorted DP CD3low thymocytes from WT mice were
analyzed. (C-D) Thymocytes from CD8WT and CD8.4 OTI Rag2−/− β2m−/− mice were
analyzed. Representative histograms (A, C) and aggregate data (B, D) (mean ± SD, n = 3-5)
are shown. P values were calculated using Student's T test (2 tailed, unequal variance). See
also Figure S2. (E) Lck was immunoprecipitated from lysates from nontreated (NT),
Cell. Author manuscript; available in PMC 2015 October 09.
Stepanek et al.
Page 21
NIH-PA Author Manuscript
pervanadate (PV), or 20 μM PP2 treated CD8WT and CD8.4 OTI Rag2−/− β2m−/−
thymocytes. Phosphorylation of Lck was analyzed by Western blotting using simultaneous
staining with Abs specific for phosphorylated or non-phosphorylated Y394. The membrane
was re-probed with Ab to total Lck. Percentage of phosphorylated Lck molecules in resting
CD8WT or CD8.4 DP thymocytes was calculated. CD8WT: n=4; CD8.4: n=5.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Cell. Author manuscript; available in PMC 2015 October 09.
Stepanek et al.
Page 22
NIH-PA Author Manuscript
Figure 3. Enhanced Lck coupling lowers the threshold for negative selection
NIH-PA Author Manuscript
Fetal thymi from CD8WT and CD8.4 OT-I Rag2−/−β2m−/− mice were exposed to OVA
derived APLs at the indicated concentrations Percentage of CD8β+ single positive cells
versus 1/potency of the ligands (Daniels et al., 2006) is shown (mean ± SEM, n=2-7). The
squares (A) show percentage of single positive cells generated with no peptide (mean ± SD,
n = 11-12). The threshold for negative selection is marked by dashed vertical lines. Student's
T test (2 tailed, unequal variance): *p<0.05, ** p<0.01, ***<0.001. See also Table S1.
NIH-PA Author Manuscript
Cell. Author manuscript; available in PMC 2015 October 09.
Stepanek et al.
Page 23
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Figure 4. CD8.4 enhances proximal signaling in OT-I thymocytes
(A-G) Thymocytes from CD8WT or CD8.4 OT-I Rag2−/−β2m−/− mice were stimulated with
100 nM Kb-OVA, -Q4R7, or -Q4H7 tetramers or left unstimulated (ns). Results were
normalized to the average of signal from unstimulated CD8WT or CD8.4 cells in each
experiment. (A) Phosphorylation of TCRζ (Y142) was analyzed by flow cytometry on
CD4+CD8+ population. Mean ± SEM, n=6. (B-C, G) Phosphorylation of LAT, ZAP70, and
Erk was analyzed using phospho-specific Abs. Reprobing the membranes for total ZAP70,
LAT, and Erk served as respective loading controls. (D-F) Phosphorylation of LAT, SLP76,
Cell. Author manuscript; available in PMC 2015 October 09.
Stepanek et al.
Page 24
NIH-PA Author Manuscript
and VAV in whole cell lysates was determined using anti-pTyr Ab and anti-actin Ab. as a
loading control by Western blotting. Mean ± SEM, n=4. Statistical significance was tested
using student's T test (1 tailed, unequal variance): * p ≤ 0.05, ** p <0.01. See also Figure S3.
(H) CD8WT or CD8.4 OT-I DP thymocytes were loaded with Indo-1 and stimulated with
200 nM Kb-OVA, -Q4R7, or -Q4H7 tetramers or 1.5 μM ionomycine. Calcium mobilization
was analyzed by flow cytometry. Ca2+ response index is shown (see Extended Matherial and
Methods). A representative experiment from a total of 3 is shown.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Cell. Author manuscript; available in PMC 2015 October 09.
Stepanek et al.
Page 25
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Figure 5. CD8.4 preferentially enhances response to weak ligands
NIH-PA Author Manuscript
Thymocytes from CD8WT and CD8.4 OT-I Rag2−/−β2m−/− mice were incubated with APCs
loaded with varying concentrations of different peptides. After 24 hours, the percentage of
CD69+ thymocytes was measured by flow cytometry. Response to OVA (A), T4 (B), and
G4 (C) is shown. Mean ± SEM, n=4. Statistical significance was tested using Student's T
test (1 tailed, unequal variance): * p<0.05, ** p<0.01. (D) CD69 response of CD8WT and
CD8.4 OT-I DP thymocytes to different antigens was examined and EC50 values were
calculated. Ratio of EC50 CD8WT/EC50 CD8.4 was plotted vs. EC50 CD8WT. Results
show that CD8.4 thymocytes are preferentially more sensitive to weaker ligands, compared
to CD8WT thymocytes. Red line shows the log-log line fit and the black lines represent 95%
confidence intervals. See also Figure S4.
Cell. Author manuscript; available in PMC 2015 October 09.
Stepanek et al.
Page 26
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Figure 6. Markov chain model for Lck delivery to the TCR-pMHC pair
NIH-PA Author Manuscript
(A) Scheme of the Markov chain model describing kinetics of protein interactions in the
plasma membrane. TM is a TCR-pMHC pair, C is an empty coreceptor, and LC is an Lck
coupled coreceptor. TM+C and TM+LC represent coreceptor and TCR-pMHC pairs (in
close proximity), TM:C and TM:LC represent coreceptor-TCR-pMHC complexes (binding).
The complex of Lck-coupled coreceptor and TCR-pMHC (TM:LC) is an absorbing end
state. KD, kb, ku, kf0, kf1 represent kinetic rates (Table S2). The rates that depend on the
extent of coreceptor-Lck coupling are shown in red. (B) Numerical solution of Markov
chain model. The probability of TM:LC complex formation as a function of time was
calculated for CD4 (pMHCII), CD8WT (pMHCI), and CD8.4 (pMHCI). Both the
probability of recruitment of any Lck-coupled coreceptor (solid line) or a coreceptor coupled
to active Lck (pY394) is shown (dashed line). (C) Probability of TM:LC pair formation as a
function of CD8-Lck coupling for MHCI ligand. The values for CD8WT (blue dot) and
CD8.4 (red dot) are marked. Other parameters remained fixed. (D) The probabilities of
TM:LC formation as a function of lattice spacing, coreceptors number, CD8 and TCR
diffusion coefficient and CD8:MHC unbinding ku (when ku/kb ratio was fixed) for MHCI
Cell. Author manuscript; available in PMC 2015 October 09.
Stepanek et al.
Page 27
ligand and CD8 coreceptor. The original parameters are marked (black dot). See also
Extended Material and Methods, Figure S5, and Table S2.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Cell. Author manuscript; available in PMC 2015 October 09.
Stepanek et al.
Page 28
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Figure 7. Model of Lck recruitment combined with kinetic proofreading predicts experimental
Graphs show TCR signal intensity as a function of number of cognate ligands at the
thymocyte/APC interface for different ligands (kon=0.1 μm2s−1, τ1/2 variable). Horizontal
green dashed line shows the likely threshold when 2 TCRs are triggered and still occupied.
(A) ‘Lck come&stay/signal duration’ model for CD8WT OT-I Rag2−/− β2m−/− thymocytes.
N. negative selectors, T. threshold ligands (partial negative selectors), P. positive selectors.
(B) Pure TCR occupancy model for OT-I Rag2−/− β2m−/− thymocytes. (C) Lck come&stay/
Cell. Author manuscript; available in PMC 2015 October 09.
Stepanek et al.
Page 29
NIH-PA Author Manuscript
signal duration model for CD8.4 OT-I Rag2−/− β2m−/− thymocytes. (D) Difference between
CD8.4- and CD8WT-mediated TCR signaling (CD8.4/CD8WT ratio) in OT-I Rag2−/−
β2m−/− thymocytes vs τ1/2 of a TCR ligand, predicted by the ‘Lck come&stay/signal
duration’ model. Advantage of increased Lck coupling (CD8.4) is increasingly apparent at
low τ1/2 (E) Comparison of TCR responses induced by pMHCI and pMHCII ligands in
polyclonal preselection DP thymocytes predicted by the Lck come&stay/signal duration
model. See also Figure S6, Table S3 and Extended Material and Methods.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Cell. Author manuscript; available in PMC 2015 October 09.

Coreceptor Scanning by the T Cell Receptor Provides a Please share